In fact, it’s useful to think of a softmax output layer with log-likelihood cost as being quite similar to a sigmoid output layer with cross-entropy cost。事实上这两者是一致的,logistic回归用的就是sigmoid函数,softmax回归是logistic回归的多类别推广。log-likelihood代价函数在二类别时就能够化简为交叉熵代价函数的形式。详细能够參考UFLDL教程
欢迎光临 电子技术论坛_中国专业的电子工程师学习交流社区-中电网技术论坛 (http://bbs.eccn.com/) | Powered by Discuz! 7.0.0 |